sign in sign up
<<
Home , Irritability, Spectrum disorder

Molecular (2010) 15, 856–867 & 2010 Macmillan Publishers Limited All rights reserved 1359-4184/10 www.nature.com/mp ORIGINAL ARTICLE The importance of irritability as a symptom of major depressive disorder: results from the National Comorbidity Survey Replication M Fava1, I Hwang2, AJ Rush3,4, N Sampson2, EE Walters2 and RC Kessler2 1Department of Psychiatry, Massachusetts General Hospital, Boston, MA, USA; 2Department of Health Care Policy, Harvard Medical School, Boston, MA, USA; 3Department of Psychiatry, University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA and 4Department of Clinical Sciences, University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA

Irritability is a diagnostic symptom of major depressive disorder (MDD) in children and adolescents but not in adults in both the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) and International Classification of Diseases (ICD-10) systems. We explore the importance of irritability for subtyping adult DSM-IV MDD in the National Comorbidity Survey Replication (NCS-R), a national US adult household survey. The WHO Composite International Diagnostic Interview (CIDI) was used to assess prevalence of many DSM-IV disorders in the lifetime and in the year before interview (12-month prevalence). MDD was assessed conventionally (that is, requiring either persistent or loss of ), but with irritability included as one of the Criterion A symptoms. We also considered the possibility that irritability might be a diagnostic symptom of adult MDD (that is, detect cases who had neither sad mood nor loss of interest). Twelve-month MDD symptom severity was assessed with the Quick Inventory of Depressive Symptomatology and role impairment with the Sheehan Disability Scale. After excluding bipolar disorders, irritability during depressive episodes was reported by roughly half of respondents with lifetime DSM-IV MDD. Irritability in the absence of either sad mood or loss of interest, in comparison, was rare. Irritability in MDD was associated with early age of onset, lifetime persistence, comorbidity with and impulse-control disorders, fatigue and self-reproach during episodes, and disability. Irritability was especially common in MDD among respondents in the age range 18–44 and students. Further investigation is warranted of distinct family aggregation, risk factors and treatment response. Consideration should also be given to including irritability as a nondiagnostic symptom of adult MDD in DSM-V and ICD-11. Molecular Psychiatry (2010) 15, 856–867; doi:10.1038/mp.2009.20; published online 10 March 2009 Keywords: epidemiology; irritability; major depressive disorder; National Comorbidity Survey Replication; adult

Introduction fact that irritability is commonly found in clinical samples of adults with MDD.3–5 The clinical literature Clinical studies of depressed children and adoles- also suggests that irritability might be a meaningful cents have shown that the most frequently reported subtyping variable in MDD, with irritable cases more 1,2 symptom in moderate is irritability, likely than nonirritable cases to be women, young, which is consistent with the Diagnostic and Statis- unemployed, more severely depressed, lower in tical Manual of Mental Disorders, fourth edition functional status and quality of life, and to have a (DSM-IV) stipulation that irritability is a diagnostic history of at least one suicide attempt.4 These symptom of major depression in children and differences could be of considerable importance, adolescents (that is, it detects subjects not detected because irritability with attacks might be by sad mood or loss of interest). However, DSM-IV present in more than one-third of patients with does not include irritability as a symptom of major MDD,6–8 although the robustness of these results is depressive disorder (MDD) among adults, despite the difficult to assess due to the fact that irritability was not assessed consistently in these studies. Correspondence: Dr R Kessler, Department of Health Care Policy, The above results all come from clinical samples. Harvard Medical School, 180 Longwood Ave, Boston, MA 2115, No study, to our knowledge, has investigated the USA. prevalence or correlates of irritability as a symptom of E-mail: [email protected] Received 24 June 2008; revised 27 January 2009; accepted 28 MDD in a general population sample. The current January 2009; published online 10 March 2009 report presents such results from the National Irritability as a symptom of MDD M Fava et al 857 Comorbidity Survey Replication (NCS-R),9 with the the level. The CIDI assessment of major goal of exploring the importance of irritability as a depressive episode (MDE) asked about symptoms in symptomatic subtyping distinction in a general the worst lifetime episode and included a number of population sample of people with a lifetime history symptoms in addition to those specified in DSM-IV. of MDD. One of these was irritability, which was assessed with a simple yes–no question about whether the respon- dent was ‘irritable, grouchy or in a bad mood’ most Materials and methods every day during the worst 2 weeks of the index Sample episode. In retrospect, the decision to assess irrit- The NCS-R is a nationally representative survey of ability with only a single question was unfortunate, as mental disorders among English-speaking household we could have evaluated the sensitivity of results to residents ages 18 and older in the continental United different definitions if multiple items had been used. States.9 Interviews were conducted with 9282 respon- Furthermore, inclusion of the term ‘bad mood’ in the dents between February 2001 and April 2003. Verbal irritability question might have led to some false informed consent was obtained before data collection. positives, to the extent that respondents interpreted Consent was verbal rather than written to maintain this term to mean a sad mood, but our impression consistency with the baseline NCS.10 The Human from subsequent debriefing interviews is that this was Subjects Committees of Harvard Medical School and not a common interpretation in the context of the the University of Michigan both approved the earlier terms ‘irritable’ and ‘grouchy’. Other symp- recruitment and consent procedures. Respondents toms included in the assessment were , were paid $50 for participation. The response rate extreme irritability and several other symptoms of was 70.9%. A probability subsample of respondents – that were included to distinguish that oversampled Composite International Diagnostic depressive episodes from mixed episodes. CIDI–SCID Interview (CIDI) cases was administered the lifetime concordance was found to be good for lifetime nonpatient version of the Structured Clinical Inter- diagnoses of MDE (k = 0.59) and excellent for diag- view for DSM-IV (SCID)11 to validate CIDI diagnoses. noses of BP spectrum disorder, including BP-I, BP-II These clinical reappraisal study respondents were and subthreshold BP disorder (k = 0.94). given a $50 incentive. A probability subsample of In addition to diagnosing threshold MDD, the CIDI nonrespondents was administered a brief telephone included an assessment of subthreshold cases, mak- survey and results were used to weight the main ing it possible to calculate how much the estimated sample for nonresponse bias. Nonrespondent survey prevalence of MDD would increase, if irritability participants were given a $100 incentive. could substitute for sad mood if the latter was not The main NCS-R interview was administered in present. Furthermore, in an effort to explore the two parts. Part I included a core diagnostic assess- implications of including irritability as a core symp- ment of all respondents (n = 9282). Part II included tom of MDD (that is, as a symptom that could questions about correlates and additional disorders substitute for the requirement of either sad mood or administered to all Part I respondents who met loss of interest if neither of the latter was present) lifetime criteria for any core disorder plus a roughly among adults rather than only among children and one-in-three probability subsample of other respon- adolescents, a separate assessment of all other dents (n = 5692). A more detailed discussion of NCS-R symptoms of MDD was made for episodes of irrit- sampling and weighting procedures is presented ability in the absence of either depressed mood or loss elsewhere.12 of interest. Respondents who met DSM-IV/CIDI criteria life- Diagnostic assessment time MDD and reported having an episode in the year Diagnoses were based on Version 3.0 of the World before the interview (12-month MDD) were adminis- Health Organization CIDI,13 a fully structured instru- tered the Quick Inventory of Depressive Symptoma- ment designed for use by trained lay interviewers tology Self-Report (QIDS-SR)15 to assess symptom who do not have clinical experience. Diagnoses are severity in the worst month of the past year. The based on DSM-IV criteria.14 Diagnostic hierarchy QIDS-SR is a fully structured measure that is strongly rules were applied in making diagnoses. The core related both to the clinician-administered Inventory disorders assessed in the survey include mood of Depressive Symptomatology (IDS-C)16 and to the disorders (MDD, dysthymia and bipolar (BP) I and II Hamilton Rating Scale of Depression (HRSD).17 disorders), anxiety disorders ( disorder, agora- Transformation rules developed for the QIDS-SR18 phobia without panic, specific phobia, social phobia, were used to convert scores into clinical severity generalized anxiety disorder (GAD), obsessive–com- categories mapped to conventional HRSD ranges of pulsive disorder, posttraumatic stress disorder none (that is, not clinically depressed), mild, moder- (PTSD)), substance disorders (alcohol and drug abuse ate, severe and very severe. These respondents were and dependence) and impulse-control disorders also administered the Sheehan Disability Scales (oppositional-defiant disorder, conduct disorder, in- (SDS)19 to assess the extent to which depression termittent explosive disorder (IED)). Age of onset interfered with functioning in work, household, (AOO) was assessed with retrospective self-reports at relationship and social functions in the worst month

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 858 of the past year. Responses were scored with a 0–10 Table 1 Disaggregated lifetime prevalence estimates of visual analogue scale having response options labeled DSM-IV/CIDI MDE in the NCS-R (n = 9282) none (0), mild (1–3), moderate (4–6), severe (7–9) and a very severe (10). Prevalence Proportion of all MDE Analysis methods Subgroup comparisons were used to study prevalence % (s.e.) % (s.e.) and correlates of lifetime irritable and nonirritable MDD. Sociodemographic correlates were examined Bipolar I–II 1.8 (0.2) 9.2 (0.9) using logistic regression. Age of onset (AOO) distribu- Subthreshold BP disorder 0.8 (0.1) 4.2 (0.5) tions were estimated using the two-part actuarial Core hypomanic symptoms 6.3 (0.3) 33.1 (1.3) method implemented in SAS.20 Persistence was Non-BP irritable MDE 5.3 (0.2) 27.7 (1.0) Non-BP non-irritable MDE 5.0 (0.2) 25.8 (1.0) examined by calculating means of reported years in Total MDE 19.2 (0.5) any lifetime depressive episode and the proportion of (n)b (9282) (1829) lifetime cases who were in an episode in the past 12 months. Twelve-month clinical severity and severity of role impairment were examined by calculating dis- Abbreviations: BP, bipolar; MDE, major depressive episode. tributions within the irritable and nonirritable depres- aPrevalence estimates are reported on the base of the entire sion subgroups. Lifetime comorbidity was assessed by sample. For example, the 1.8% prevalence of MDE with BP calculating odds ratios (ORs) of lifetime irritable and disorder I–II means that 1.8% of the sample both met nonirritable MDD with other lifetime DSM-IV/CIDI lifetime criteria for MDE and lifetime criteria for BP disorder disorders. Because the NCS-R sample design used I–II. b weighting and clustering, all statistical analyses were The reported sample sizes are unweighted and assessed in carried out using the Taylor series linearization the part I sample. method,21 a design-based method implemented in the SUDAAN software system.22 Significance tests of sets long as they were accompanied by at least four other of coefficients were made using Wald w2-tests based on DSM-IV Criterion A symptoms and all other DSM-IV design-corrected coefficient variance–covariance criteria, lifetime prevalence of irritable MDD would matrices. Statistical significance was evaluated using increase by another 0.5% (0.1) to 6.4% (0.6). To put this two-sided design-based 0.05 level tests. last point in perspective, only 28 additional people out of a sample of 9282 would meet lifetime criteria for Results MDE and not for major depression if we allowed irritability to be a diagnostic symptom of adult MDD. Prevalence, age of onset and persistence Mean retrospectively reported AOO of DSM-IV Of the 19.2% of NCS-R respondents who met lifetime MDD (that is, excluding the two kinds of exploratory criteria for MDE, roughly one-eighth (13.4% of cases described in the previous paragraph, neither of the 19.2%) were classified as having either threshold which meets DSM-IV criteria) was found to be or subthreshold BP disorder, and another 33.1% significantly earlier for irritable (26.7 years) than

reported a lifetime history of core hypomanic nonirritable (31.3 years) cases (F1,953 = 13.7, P < 0.001; symptoms (that is, a distinct period of abnormally Table 2). The shape of the AOO distribution is and persistently elevated, expansive or irritable nonetheless generally consistent for the two subtypes mood lasting several days or longer with at least one (Figure 1). The estimated mean number of years in other symptom of hypomania) (Table 1). Roughly any lifetime episode is also similar for irritable and

equal numbers of the remaining 53.5% of respon- nonirritable MDD (5.7 vs 5.1, F1,953 = 0.1, P = 0.75), dents with lifetime MDE reported either the pres- although persistence, as indirectly indicated by the ence (27.7%; n = 489) or the absence (25.8%; n = 466) ratio of 12-month prevalence to lifetime prevalence, is of irritability in their worst lifetime depressive significantly higher for respondents with irritable 2 episode, representing 5.3% (irritable MDE) and 5.0% (40.3%) than nonirritable (28.8%) MDD (w1 = 9.0, (nonirritable MDE), respectively, of the total NCS-R P = 0.004). sample. The estimated lifetime prevalence (with standard Symptom profiles and episode severity error in parentheses) of irritable MDD would increase Tetrachoric factor analysis of the nine DSM-IV by 0.6% (0.1), from 5.3% (0.2) to 5.9% (0.2), if Criterion A symptoms of MDD and irritability among irritability was included as a 10th Criterion A NCS-R respondents who endorsed a CIDI diagnostic symptom and if diagnosis required sad mood or loss stem question for lifetime major depression (that is, a of interest along with a total of at least 5 of 10 period lasting 2 weeks or longer when the respondent Criterion A symptoms (rather than the requirement of experienced either sad mood most of the day nearly at least 5 of 9 in DSM-IV). If irritability was included every day or loss of interest) was carried out. A strong as a diagnostic symptom of adult MDD, so that first principal factor (with a 0.85 cumulative propor- episodes of irritability in the absence of either sad tion of eigenvalues compared to 0.20 for the second mood or loss of interest qualified for a diagnosis so unrotated principal factor) was found in which the

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 859 Table 2 Onset and course of irritable and nonirritable DSM- P = 0.10), with 48.1% of irritable and 45.0% of IV/CIDI MDD nonirritable cases classified either severe or very severe (Table 4). Mean scores on the SDS also are Irritable Nonirritable quite similar for the two groups, although there is a MDE consistent trend for disability to be slightly higher for irritable than nonirritable MDD (Table 5). Est (s.e.) Est (s.e.) F/w2 Sociodemographic correlates Mean age of onset 26.7a (0.7) 31.3 (0.9) 13.7a,b b The sociodemographic correlates of irritable MDD Mean years in episode 5.7 (0.5) 5.1 (0.9) 0.1 are generally similar to those of nonirritable MDD 12-month: lifetime 40.3a (2.7) 28.8 (1.6) 9.0a,c prevalence (Table 6). Both are significantly more common among (n)d (497) (480) women than men, with statistically indistinguishable 2 female/male ORs (1.6 vs 1.9, w1 = 0.6, P = 0.42). Both Abbreviations: Est, ; MDE, major depressive episode. are inversely related to income and education, aSignificant difference between irritable and nonirritable although somewhat more strongly so for irritable than cases at the 0.05 level, two-sided test. nonirritable cases. Both are more prevalent among the bF-test with 1 and 953 degrees of freedom. previously married than the currently married and cw2-test with 1 degree of freedom. among non-Hispanic whites than non-Hispanic dThe reported sample sizes are unweighted and assessed in blacks or Hispanics. the part I sample. There are also notable differences in the associa- tions of irritable and nonirritable MDD with age and age-related sociodemographic variables. Irritable 100 MDD is more prevalent than nonirritable MDD among 90 respondents in the age range 18–44, students, the

80 never married and respondents in the ‘other’ race- ethnic group (that is, neither non-Hispanic white, 70 non-Hispanic black or Hispanic). Nonirritable MDD, 60 in comparison, is more prevalent than irritable MDD 50 among respondents ages 60 þ . More detailed analyses

40 (results available on request) showed that these Cumulative % differences are not due to differences in comorbid 30 Irritable depression Non-Irritable depression DSM-IV/CIDI anxiety, impulse-control or substance 20 disorders. 10

0 Comorbidity with other DSM-IV disorders 0 10 20 30 40 50 60 70 80 Lifetime comorbidity with other DSM-IV/CIDI dis- Time until onset of disorder orders was reported by a significantly higher propor- Figure 1 Age of onset for major depressive episode (MDE). tion of respondents with irritable (71.3%) than nonirritable (57.9%) MDD (F1,953 = 6.8, P = 0.013), although ORs are consistently greater than 1.0 and factor loading for irritability (0.31) is in the lower end generally significant for both subtypes (Table 7). The of the interquartile range (IQR: 25th–75th percentile) ORs of irritable MDD are higher than those of of the factor loadings of the DSM-IV symptoms (IQR: nonirritable MDD with dysthymia and seven of the 0.30–0.54). (More detailed results available on re- nine anxiety assessed in the NCS-R. Four quest.) of the latter seven are significant at the 0.05 level. It is Symptom profiles for the nine DSM-IV Criterion noteworthy that these four—GAD, panic attack, social A symptoms in the worst lifetime episode are phobia and PTSD—include the anxiety disorders quite comparable for irritable and nonirritable MDD with the latest AOO. The ORs of irritable MDD are (Table 3). Significant differences are limited to fatigue also higher than those of nonirritable MDD with all and self-reproach, both of which are significantly three of the impulse-control disorders assessed in the more prevalent in irritable than nonirritable MDD. core NCS-R: -deficit/hyperactivity disorder, Irritable cases were also marginally more likely than oppositional-defiant disorder and IED. It is note- nonirritable cases to report morbid thoughts of death worthy that all these ORs are statistically significant 2 (71.5 vs 65.5%, w1 = 3.6, P = 0.06). More detailed for irritable MDD, whereas none is significant for analyses (results available on request) showed that nonirritable MDD. IED might be thought to have an these differences were not due to differences in especially strong association with irritability, as the comorbid DSM-IV/CIDI anxiety, impulse-control or uncontrollable anger attacks of IED could be seen as substance disorders. extreme forms of irritability. Yet only 13.7% of The distribution of clinical severity, as defined respondents with irritable MDD have a lifetime by the QIDS-SR, does not differ significantly for history of IED. Although this is considerably higher 2 12-month irritable vs nonirritable MDD (w3 = 6.7, than the 5.5% lifetime prevalence of IED among

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 860 Table 3 Symptom profilesa of irritable and nonirritable DSM-IV/CIDI MDE

Irritable Nonirritable

% (s.e.) % (s.e.) w2

Sad mood 99.1 (0.5) 98.6 (0.7) 0.3 Loss of interest 88.2 (2.0) 85.0 (1.8) 1.5 Appetite of weight disturbance Appetite/weight gain 18.6 (2.3) 15.8 (2.1) 1.3 Appetite/weight loss 68.9 (2.7) 72.7 (2.5) 1.4 Sleep disturbance 16.7 (1.8) 17.5 (1.5) 0.1 78.0 (2.1) 73.9 (1.7) 2.3 Activity disturbance 8.7 (1.6) 7.3 (0.9) 0.5 Psychomotor retardation 41.8 (2.6) 37.2 (2.5) 1.9 Fatigue 89.5b (1.4) 83.5 (1.9) 11.5b Self-reproach or 81.3b (1.5) 65.9 (1.9) 27.8b Poor concentration or indecisiveness 91.3 (1.4) 87.4 (1.9) 2.3 Morbid thoughts of death 71.5 (2.8) 65.5 (2.4) 3.6 (n)c (497) (480)

aRefers to symptoms in persons with a major depressive episode in the last 12 months. bSignificant difference between irritable and nonirritable MDE at the 0.05 level, two-sided test. cThe reported sample sizes are unweighted and assessed in the part I sample.

Table 4 Clinical severity (QIDS-SR) of 12-montha irritable Table 5 Role disability (Mean SDS score) of 12-montha and nonirritable DSM-IV/CIDI MDD irritable and nonirritable DSM-IV/CIDI MDE

Irritable* Nonirritable Irritable Nonirritable

% (s.e.) % (s.e.) Mean (s.e.) Mean (s.e.) F1,339

Mild 10.4 (2.0) 15.7 (3.0) Home 5.2 (0.2) 5.0 (0.3) 0.2 Moderate 41.5 (3.7) 39.3 (3.7) Work 4.4 (0.2) 4.0 (0.4) 0.9 Severe 38.2 (3.4) 28.9 (3.2) Interpersonal 4.7 (0.2) 4.2 (0.3) 3.1 Very severe 9.9 (2.6) 16.1 (2.7) Social 5.6 (0.2) 5.1 (0.3) 2.2 (n)b (202) (145) Maximum disability 6.9 (0.2) 6.4 (0.2) 2.7 (n)b (202) (145)

*Significance of difference between irritable and nonirri- aTwelve-month MDD is an episode of MDD that occurred at 2 table MDE: w3 = 6.7, P = 0.08. any time within 1 year of the interview. Role disability was aTwelve-month MDD is an episode of MDD that occurred at assessed for the most severe month in the past year. any time within 1 year of the interview. Severity was bThe reported sample sizes are unweighted and represents assessed for the most severe month in the past year. the 12-month irritable/nonirritable cases in the part I bThe reported sample sizes are unweighted and assessed in sample. the part I sample. Cases are 12-month cases instead of lifetime compared to previous tables. P = 0.06; Table 8). This difference was especially pronounced among those whose 12-month depression respondents with nonirritable MDD, it is clear that was associated with a very severe QIDS-SR score IED is a minority phenomenon among people with (84.5 vs 51.4%, z = 3.6, P < 0.001). A significantly irritable MDD. The ORs of irritable and nonirritable higher proportion of irritable than nonirritable cases MDD with substance disorders, finally, are generally were also treated by a psychiatrist, although this equivalent. difference was confined to very severe cases (84.5 vs 43.8%, z = 4.4, P < 0.001). Treatment Respondents with lifetime irritable MDD who had a Discussion 12-month depressive episode were somewhat more likely than 12-month cases with nonirritable MDD to The results reported here are limited by the fact that receive professional treatment for their depression the NCS-R used a fully structured lay-administered within a year of the interview (64.5 vs 53.5%, z = 1.9, diagnostic interview rather than a semistructured

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 861 Table 6 Sociodemographic correlates of lifetime irritable and nonirritable DSM-IV/CIDI MDE

Irritable Nonirritable Irritable: nonirritable

% (s.e.) OR (95% CI) % (s.e.) OR (95% CI) w2 d.f.

Sex Female 7.6 (0.4) 1.6a (1.3–2.1) 7.3 (0.5) 1.9a (1.5–2.4) Male 4.8 (0.5) 1.0 — 3.9 (0.4) 1.0 — 0.6 1 2 a a w1 14.5 31.9

Age 18–29 5.7 (0.5) 1.8a (1.3–2.6) 2.6 (0.4) 0.4a (0.3–0.6) 30–44 8.2 (0.7) 2.7a (1.8–3.9) 5.3 (0.5) 1.0 (0.7–1.3) 45–59 7.2 (0.6) 2.3a (1.6–3.4) 8.9 (0.8) 1.6a (1.2–2.3) 60 þ 3.2 (0.5) 1.0 — 5.6 (0.7) 1.0 — 47.8b 3 2 a a w3 30.4 52.7

Education 0–11 3.4 (0.6) 1.0 — 3.7 (0.7) 1.0 — 12 6.0 (0.5) 1.8a (1.2–2.6) 5.2 (0.4) 1.5 (0.9–2.2) 13–15 6.8 (0.6) 2.1a (1.4–3.1) 6.0 (0.6) 1.7a (1.1–2.5) 16 þ 7.8 (0.8) 2.4a (1.6–3.7) 7.5 (0.6) 2.2a (1.4–3.3) 17.0b 3 2 a a w3 17.7 26.7

Employment status Work 6.7 (0.4) 1.0 — 6.1 (0.4) 1.0 — Student 6.4 (2.1) 0.9 (0.4–2.1) 0.9 (0.7) 0.1a (0.0–0.6) Homemaker 5.8 (1.1) 0.9 (0.6–1.3) 5.1 (1.1) 0.8 (0.5–1.4) Retired 3.6 (0.6) 0.5a (0.4–0.7) 5.4 (0.9) 0.9 (0.6–1.3) Other 7.4 (1.6) 1.1 (0.7–1.8) 5.4 (0.8) 0.9 (0.6–1.2) 7.9 4 2 a w4 14.8 8.0

Family incomed Low 5.4 (0.6) 1.0 — 4.6 (0.6) 1.0 — Low average 5.1 (0.5) 0.9 (0.6–1.3) 4.9 (0.7) 1.1 (0.7–1.4) High average 6.5 (0.5) 1.2 (0.7–1.2) 6.2 (0.5) 1.4a (1.1–2.0) High 7.6 (0.7) 1.5 (0.9–1.8) 6.6 (0.5) 1.5a (1.1–2.1) 0.5 3 2 a a w3 10.2 14.2

Marital status Married/ 5.3 (0.4) 1.0 — 5.6 (0.4) 1.0 — cohabitating Separated/ 8.2 (0.7) 1.6a (1.3–2.0) 9.1 (1.0) 1.7 (1.3–2.2) widowed/ divorced Never 6.8 (0.7) 1.3a (1.0–1.7) 2.9 (0.5) 0.5a (0.3–0.8) 2.6 2 married 2 a a w2 21.9 56.7

Race/ethnicity Hispanic 3.9 (0.7) 0.5a (0.4–0.8) 4.5 (0.8) 0.7a (0.4–1.0) Black 3.4 (0.6) 0.5a (0.3–0.7) 2.2 (0.6) 0.3a (0.2–0.6) Other 8.2 (1.4) 1.1 (0.8–1.6) 4.3 (1.0) 0.6a (0.4–0.9) White 7.0 (0.4) 1.0 — 6.5 (0.4) 1.0 — 4.3 3 2 a a w3 25.5 19.8

Region Northeast 6.6 (0.8) 1.0 — 5.6 (0.9) 1.0 — South 6.3 (0.7) 0.9 (0.7–1.3) 5.6 (0.6) 1.0 (0.7–1.5) Midwest 5.7 (0.5) 0.9 (0.6–1.2) 5.0 (0.6) 0.9 (0.6–1.3) West 6.6 (0.6) 1.0 (0.7–1.4) 7.0 (0.7) 1.3 (0.8–1.9) 3.0 3 2 w3 1.9 5.6 (n)c (497) (7950) (480) (7933) aSignificant at the 0.05 level, two-sided test. bSignificant difference between irritable and nonirritable MDE at the 0.05 level, two-sided test. cThe reported sample sizes are unweighted. Prevalences represent the cases with irritable/nonirritable MDE among the demographic categories, whereas the models are assessed among cases without lifetime MDE in addition to people with either irritable or nonirritable MDE. There are 7453 cases in the part I sample without lifetime MDE, and 497/480 cases for irritable/nonirritable, respectively. Thus, the models for irritable MDE are in a subsample of 7453 þ 497, whereas for nonirritable MDE, it is 7453 þ 480. dIncome was assessed in the part II sample. Models assessed among part II cases without Lifetime MDE and part II cases with irritable/nonirritable MDE.

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 862 Table 7 Lifetime comorbidity of irritable and nonirritable DSM-IV/CIDI MDE with other DSM-IV/CIDI disorders

Irritable Nonirritable Irritable: nonirritable 2 d % (s.e.) OR (95% CI) % (s.e.) OR (95% CI) w1

Mood disorders Dysthymia 17.0 1.6 47.6a (25.4–89.5) 11.8 2.1 27.3a (14.1–52.9) 5.6b

Anxiety disorders GAD 25.7 1.8 9.2a (6.9–12.3) 16.4 1.9 4.8a (3.4–6.7) 15.7b Panic attack 47.4 2.2 3.0a (2.4–3.8) 37.8 2.7 2.1a (1.7–2.7) 8.7b 9.3 1.1 3.0a (2.1–4.1) 8.2 1.5 2.6a (1.6–4.0) 0.2 Social phobia 23.6 1.7 3.4a (2.8–4.1) 17.0 1.7 2.4a (1.9–3.1) 6.8b Specific phobia 20.6 1.7 2.3a (1.8–3.0) 17.3 1.8 1.9a (1.5–2.4) 1.7 3.5 1.0 2.7a (1.5–4.9) 3.9 1.0 3.1a (1.7–5.6) 0.1 Posttraumatic stress 17.7 1.9 4.6a (3.2–6.5) 11.3 1.4 2.8a (1.9–4.0) 5.2b disordere Separation anxiety disordere 5.9 1.0 1.9a (1.3–2.8) 7.0 2.0 2.6a (1.6–4.4) 1.1 Obsessive-compulsive 2.2 0.7 7.5a (2.5–22.7) 0.5 0.5 1.5 (0.2–12.2) 2.3 disorderf Any anxiety disordere 54.4 1.7 4.4a (3.6–5.3) 44.8 2.9 3.1a (2.5–3.9) 11.1b

Impulse-control disorders Attention-deficit/ 6.6 1.4 2.4a (1.3–4.3) 3.6 1.6 1.4 (0.6–3.4) 1.8 hyperactivity disorderg Oppositional defiant 7.8 1.2 1.9a (1.1–3.2) 4.8 1.6 1.5 (0.8–2.5) 0.4 disorderg Intermittent explosive 13.7 1.7 2.7a (2.0–3.8) 5.5 1.1 1.3 (0.8–2.0) 7.0b disorder Any impulse-control 23.5 1.8 2.8a (2.1–3.8) 10.4 2.0 1.5 (0.9–2.3) 5.2b disorderg

Substance disorders Alcohol abusee 16.7 2.8 2.0a (1.2–3.2) 16.0 2.6 2.1a (1.4–3.1) 0.0 Alcohol dependencee 7.5 1.3 2.3a (1.4–3.8) 5.7 1.1 1.9a (1.3–2.9) 0.5 Drug abusee 10.4 1.8 2.0a (1.2–3.1) 10.0 1.7 2.5a (1.6–3.9) 0.5 Drug dependencee 2.6 0.7 1.4a (0.7–2.6) 4.9 1.7 3.7a (1.9–7.3) 3.9b Any substance disordere 18.1 2.8 1.9a (1.2–2.9) 17.8 2.5 2.1a (1.5–3.0) 0.1

Any disordere 71.3 2.0 5.2a (4.1–6.5) 57.9 2.9 3.3a (2.6–4.2) 6.8b (n)c (497) (7950) (480) (7933)

aSignificant at the 0.05 level, two-sided test. bSignificant difference between irritable and nonirritable MDE at the 0.05 level, two-sided test. cThe reported sample sizes are unweighted. Prevalences represent the cases with the DSM-IV diagnosis among cases with irritable/nonirritable MDE, while the models are assessed among cases without Lifetime MDE in addition to people with either irritable or nonirritable MDE. There are 7453 cases in the part I sample without lifetime MDE, and 497/480 cases for irritable/nonirritable, respectively. Thus, the models for irritable MDE are in a subsample of 7453 þ 497, whereas for nonirritable MDE, it is 7453 þ 480. dThe w2 values are for the significance of the difference between irritable and nonirritable MDD. eAssessed in the part II sample. Models assessed among part II cases without lifetime MDE and part II cases with irritable/ nonirritable MDE. fAssessed in a random one-third of the part II sample. gAssessed in the part II sample among respondents in the age range 18–44.

clinician-administered diagnostic interview. How- episode, possibly leading to more imprecision in the ever, the fact that good concordance was found distinction between irritable and nonirritable cases between CIDI diagnoses with blinded SCID clinical than if the assessment had been made using a more diagnoses reduces this concern. In addition, irritabil- detailed assessment across a number of episodes. The ity was defined as being ‘irritable, grouchy or in a bad base rate of ‘irritability’ was not assessed in the entire mood,’ and this definition may be overinclusive. sample, making it impossible to determine whether Another limitation is that irritability was assessed the high prevalence of irritability in MDD is different with only a single question for a single worst lifetime from the lifetime prevalence of irritability in the

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 863 Table 8 Treatment of 12-montha irritable and nonirritable DSM-IV/CIDI MDE by clinical severity and sector of treatment

Any treatmentb Treatment by a psychiatrist Sample sizesc

Irritable Nonirritable Irritable Nonirritable Irritable Nonirritable % (s.e.) % (s.e.) % (s.e.) % (s.e.)

Mild 49.4 (4.6) 51.3 (4.7) 16.5 (6.7) 25.4 (6.1) (46) (42) Moderate 68.5d (3.5) 53.6 (5.6) 17.4 (4.2) 21.5 (3.6) (72) (46) Severe 66.5 (7.6) 56.7 (6.7) 25.2 (6.2) 26.6 (2.2) (68) (35) Very severe 84.5d (0.1) 51.4 (8.5) 84.5d (0.1) 43.8 (9.2) (15) (17) Total 64.5 (3.6) 53.5 (4.5) 24.7 (2.8) 26.8 (3.5) (201) (140) (n) (201) (140) (201) (140) aTwelve-month MDD is an episode of MDD that occurred at any time within 1 year of the interview. bAny treatment includes treatment by a psychiatrist, another mental health professional, a general medical professional, a human services professional (e.g., spiritual advisor, social worker in a social services agency) and in the complementary- alternative medical sector (e.g., therapist, self-help group). cThe reported sample sizes are unweighted and assessed in the part II sample. dSignificant difference between irritable and nonirritable MDE at the 0.05 level, two-sided test. general population. However, the fact that irritability sion. The results were essentially unchanged (results was found to have a significant factor loading with the available upon request). The only difference was that symptoms of MDE among respondents who endorsed the rate of comorbid PTSD in irritable depression a lifetime history of either dysphoria or (16.6%) was no longer significantly higher than in shows that irritability is associated with the symp- nonirritable depression (11.3%). toms of DSM-IV major depression. This finding is Within the context of the above limitations, our consistent with clinical evidence that irritability is results provide the first nationally representative US more common among depressed vs nondepressed general population prevalence estimates of irritable vs subjects.23 nonirritable MDD. Irritability was found in roughly The above limitations are likely to lead to the half the cases of MDD, making it at least as common as results regarding differences between irritable and a number of the DSM-IV Criterion A symptoms of nonirritable cases being conservative. An additional MDD, including all the reverse vegetative symptoms limitation, though, might have more complex effects: combined (weight gain, hypersomnia, psychomotor that respondents were not followed over time or agitation) and psychomotor retardation. This finding assessed for a family history of BP disorder, raising is consistent with previous evidence that irritability is the possibility that at least some of the respondents more common among depressed than nondepressed classified as having irritable MDD might actually be subjects.23 The prevalence of irritability found here is in the BP spectrum and might eventually convert to somewhat higher than the 37–40% irritability pre- having BP disorder. On the basis of this limitation, the valence estimates reported in previous studies of claim that the cases of irritable MDD seen here are depressed outpatients.4,5 Recall, though, that these non-BP should be seen as provisional. clinical studies considered only symptoms in the A final noteworthy limitation is that the analysis current episode whereas we considered symptoms in was carried out exclusively among respondents who the worst lifetime episode. In addition, there was no met DSM-IV criteria for major depression. This means consistency in any of these studies in the methods that respondents with irritable depression were used to assess irritability. required to have six symptoms compared to five for We excluded respondents with a lifetime history of respondents with nonirritable depression. It could be BP disorder as well as those with a history of either argued that not allowing irritability to be an alter- subthreshold BP disorder24 or core hypomanic symp- native to sadness or lack of interest might have toms (that is, a distinct period of abnormally and introduced a systematic bias toward a relatively more persistently elevated, expansive or irritable mood severe form of depression in our analysis. As noted lasting several days or longer with at least one other above, though, only 28 additional cases would be symptom of hypomania) in recognition of the fact that classified as having irritable depression if we allowed depression with anger has been conceptualized by irritability to be a diagnostic symptom of MDD. This some commentators as a BP spectrum disorder.25 As means that any bias introduced in our analysis by noted above, the fact that many of the NCS-R requiring respondents with irritable depression to respondents are still in the age of risk of onset of BP have at least six symptoms would likely be modest. disorder means that some unknown number of those This was confirmed by replicating all with irritable MDD might have their diagnoses analyses reported above with these 28 respondents changed to BP disorder in the future. However, given added to those considered to have irritable depres- the very high prevalence of irritability among respon-

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 864 dents with lifetime MDD after excluding respondents conceivable that the comparatively early AOO and in the BP spectrum, and taking into consideration that young age at interview of respondents with irritable upper-bound estimates put lifetime prevalence of BP MDE are related to the fact that irritability is such a spectrum disorder at no more than 5–7% of the commonly occurring symptom in child and adoles- population,26,27 it is very likely that this sort of cent depression.1,2 Another possibility is that irrit- conversion will occur to a high proportion of the ability is becoming a more common feature of NCS-R respondents classified as having irritable MDD. depressive episodes in recent cohorts. On the basis of the above considerations, our The finding that severity of illness, as indicated by findings challenge the view that irritability in depres- the QIDS-SR and the SDS, did not differ between sion is a specific indicator of bipolarity.28,29 The irritable and nonirritable MDD diverges from the findings also raise the question whether the revised STAR*D finding that patients with irritable MDD had DSM and International Classification of Diseases greater illness severity than those with nonirritable (ICD) criteria should include irritability as a symptom MDD.4 This discrepancy may be due, at least in part, of non-BP depression. Not only did we find that to our finding that people with very severe, irritable irritability is a common feature of non-BP depression, MDD were significantly more likely to obtain psy- but we also found that irritability was meaningfully chiatric treatment than people with very severe, related to a number of significant clinical features. nonirritable MDD, leading to an overrepresentation Mean retrospectively reported AOO of MDD is of very severe cases of irritable MDD in specialty significantly earlier for irritable than nonirritable treatment samples. MDD. Although the shape of the AOO distribution NCS-R respondents with irritable MDD reported a is nonetheless generally consistent for the two higher prevalence of fatigue, self-reproach and mor- subtypes, the earlier onset of irritable than nonirri- bid thoughts of death than those with nonirritable table MDD is consistent with the one, small clinical MDD. The elevated prevalence of morbid thoughts of study of which we are aware that examined this death is indirectly consistent with the association issue.30 It is unlikely that this is due to differential between past history of suicide attempts and irritable recall bias, as post hoc analysis shows that the same MDD in the STAR*D study.4 The elevated prevalence finding holds when we compare AOO of irritable and of self-reproach in NCS-R irritable MDD might be part nonirritable MDD in subsamples that are matched on of the pattern, as self-reproach has been linked to age at interview. It is noteworthy, though, that it has .32 It is noteworthy that similar long been known that early AOO of MDE is also often symptom differences were found in the STAR*D found in BP disorder,31 again raising a question about sample, but these differences were explained by the possibility that some cases of irritable MDD are in overall illness severity.4 It is striking that this is not the BP spectrum. the case in the NCS-R, as these symptom differences Another important clinical finding is the signifi- were found despite irritable and nonirritable MDD cantly greater persistence of irritable than nonirritable not differing in overall illness severity. This suggests MDD, as defined by the proportion of lifetime cases that symptoms of self-reproach and morbid thoughts who had a 12-month depressive episode. We are aware of death, and possibly fatigue (which was not elevated of no prior study that has examined this issue. As among STAR*R patients with irritable MDD) provide noted above, the greater persistence of irritable than some genuine differentiation between irritable and nonirritable MDD was still present after we controlled nonirritable MDD. Clearly, though, these differences for lifetime comorbidity. We also adjusted for differ- need to be examined in independent samples con- ences in AOO, as irritable MDD had an earlier onset trolling for overall illness severity before we can than nonirritable MDD, which might explain the more conclude that they are due to more than idiosyncratic persistent course of irritable MDD, but we found that features of a single study. the significantly higher 12-month/lifetime prevalence The significantly higher prevalence of comorbid ratio of irritable than nonirritable MDD persisted when anxiety disorders among people with irritable than this control was introduced along with controls for nonirritable MDD is in agreement with the observa- comorbidity. However, retrospectively reported mean tion that outpatients with MDD with anxious depres- number of years with any lifetime episode did not sion in the STAR*D study were significantly more differ significantly for irritable and nonirritable MDD, likely to report irritability than those without anxious demonstrating inconsistency in the finding regarding features.33 As no previous study investigated comor- persistence depending on the indicator used to define bidity of irritable vs nonirritable MDD with impulse- persistence. As course of illness is a critical clinical control disorders, our finding that impulse-control feature, more definitive data are needed on differences disorders are comorbid only with irritable MDD is between irritable and nonirritable MDD in this regard, unique. This is potentially important as it is the most ideally from prospective studies. dramatic difference we found in the clinical corre- The NCS-R finding that irritability is more pre- lates of irritable vs nonirritable MDD. The association valent in MDD among respondents in the age range of irritable MDD with impulse-control disorders 18–44 and among students is consistent with the might account for the thus far unexplored finding in finding of a decreasing prevalence of irritability in the Epidemiologic Catchment Area Survey that MDD with age in the STAR*D clinical sample.4 It is depression was related to violent behavior.34

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 865 The findings that irritability is a common symptom Thorough evaluation of the extent to which these of MDD and that irritable MDD is associated with additional cases would have similar treatment re- distinctive clinical features raises the possibility that sponse and correlates to cases that meet current DSM- the presence of irritability might be a useful MDD IV criteria for irritable MDD is beyond the scope of subtyping distinction. Whether this turns out to be this report. However, the apparent importance of the case, though, will require further investigation of irritability in MDD documented here suggests that differential risk factors, family aggregation and treat- clinical studies are needed of the people who would ment response. A more detailed investigation of be conferred a diagnosis of MDD if irritability were family aggregation of comorbid impulse-control dis- added as a core symptom in DSM-V and ICD-11. orders might turn out to be especially illuminating in this regard in light of the unique comorbidity of Acknowledgments irritable MDD with impulse-control disorders. Finally, the association in MDD of irritability with The National Comorbidity Survey Replication specific symptom clusters and comorbidities may (NCS-R) was supported by the National Institute of reflect a psychopathological endophenotype that Mental Health (NIMH; U01-MH60220) with supple- might defy the current DSM-IV nosology and, in fact, mental support from the National Institute of Drug cut across DSM-IV clinical entities such as mood and Abuse, the and Mental Health anxiety disorders. A related example of such a cross- Services Administration, the Robert Wood Johnson cutting distinction exists in the reduction in seroto- Foundation (Grant No. 044780) and the John W Alden nergic neurotransmission, as suggested by blunted . Collaborating NCS-R investigators include prolactin responses to challenges, re- Ronald C Kessler (Principal Investigator, Harvard ported in patients with both unipolar and BP Medical School), Kathleen Merikangas (Co-principal depression,35 anxiety disorders such as PTSD36 and Investigator, NIMH), James Anthony (Michigan State patients with impulsive aggres- University), William Eaton (The Johns Hopkins sion.37 Consistent with these observations, we have University), Meyer Glantz (NIDA), Doreen Koretz found that patients with MDD, irritability and anger (Harvard University), Jane McLeod (Indiana Univer- attacks had a significantly greater blunting of the sity), Mark Olfson (Columbia University College of prolactin response to fenfluramine challenge than Physicians and Surgeons), Harold Pincus (University patients with MDD without irritability and anger of Pittsburgh), Greg Simon (Group Health Coopera- attacks.38 One could hypothesize that this particular tive), T Bedirhan Ustun (World Health Organization), endophenotype may share signs of serotonergic Michael Von Korff (Group Health Cooperative), Philip dysregulation, in addition to the clinical character- Wang (Harvard Medical School), Kenneth Wells istics that we have identified, although further studies (UCLA), Elaine Wethington (Cornell University) and would need to be carried out to assess the specificity Hans-Ulrich Wittchen (Max Planck Institute of Psy- and temporal stability of this possible subtype. chiatry). The views and opinions expressed in this With regard to differential treatment response, no report are those of the authors and should not be study has ever investigated whether irritability is a construed to represent the views of any of the moderator of antidepressant treatment response. sponsoring organizations, agencies or US Govern- However, the fact that patients with MDD with ment. A complete list of NCS publications and the irritability and anger attacks have been shown to full text of all NCS-R instruments can be found at have distinctive neurobiological features, such as http://www.hcp.med.harvard.edu/ncs. Send corre- blunting of the prolactin response to fenfluramine spondence to [email protected]. The NCS- challenge,38 and a distinctive pattern of regional R is carried out in conjunction with the World Health cerebral blood in the left ventromedial prefrontal Organization World Mental Health (WMH) Survey cortex and left amygdala during anger induction39 Initiative. We thank the staff of the WMH Data raises the possibility that there might be distinctive Collection and Data Analysis Coordination Centers responsiveness to antidepressant treatment. This is an for assistance with instrumentation, fieldwork and issue that clearly warrants additional investigation in consultation on data analysis. These activities were light of the results presented in the current report. supported by the John D and Catherine T MacArthur Even before further studies of risk factors and Foundation, the Pfizer Foundation, the US Public consequences are carried out, though, the results Health Service (R13MH066849, R01-MH069864 and reported here raise the question whether irritability R01 DA016558), Eli Lilly and Company, GlaxoS- should be included as a symptom of MDD in the mithKline, Ortho-McNeil Pharmaceutical Inc. and revised DSM-V and ICD-11 diagnostic systems. We the Pan American Health Organization. A complete saw that irritability is as strongly related as other list of WMH publications and instruments can be DSM-IV symptoms with the underlying dimension of found at (http://www.hcp.med.harvard.edu/wmhcidi). depression symptomatology in factor analysis. We saw that the prevalence of irritable MDD would be Financial Disclosures increased by roughly one-fifth (from 5.3 to 6.4% Dr Fava has research support from Abbott Labora- lifetime prevalence), but not dramatically, by includ- tories, Alkermes, Aspect Medical Systems, AstraZe- ing irritability as a core Criterion A symptom of MDD. neca, Bristol-Myers Squibb Company, Cephalon,

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 866 Eli Lilly and Company, Forest Pharmaceuticals Inc., 3 Fava GA, Grandi S, Canestrari R, Molnar G. Prodromal symptoms GlaxoSmithKline, Johnson & Johnson Pharmaceuti- in primary major depressive disorder. J Disord 1990; 19: cals, Lichtwer Pharma GmbH, Lorex Pharmaceuticals, 149–152. 4 Perlis RH, Fraguas R, Fava M, Trivedi MH, Luther JF, Wisniewski Novartis, Organon Inc., PamLab LLC, Pfizer Inc., SR et al. Prevalence and clinical correlates of irritability in major Pharmavite, Roche, Sanofi-Synthelabo, Solvay Phar- depressive disorder: a preliminary report from the Sequenced maceuticals Inc. and Wyeth-Ayerst Laboratories. He Treatment Alternatives to Relieve Depression study. J Clin has served as an advisor/consultant for Aspect Psychiatry 2005; 66: 159–166. Medical Systems, AstraZeneca, Bayer AG, Biovail 5 Snaith RP, Taylor CM. Irritability: definition, assessment and associated factors. Br J Psychiatry 1985; 147: 127–136. Pharmaceuticals Inc., BrainCells Inc., Bristol-Myers 6 Fava M, Alpert J, Nierenberg AA, Ghaemi N, O’Sullivan R, Tedlow Squibb Company, Cephalon, Comellas, Cypress Phar- J et al. Fluoxetine treatment of anger attacks: a replication study. maceuticals, DOV Pharmaceuticals, Eli Lilly and Ann Clin Psychiatry 1996; 8: 7–10. Company, EPIX Pharmaceuticals, Fabre-Kramer Phar- 7 Fava M, Nierenberg AA, Quitkin FM, Zisook S, Pearlstein T, Stone maceuticals Inc., Forest Pharmaceuticals Inc., Glaxo- A et al. A preliminary study on the efficacy of and imipramine on anger attacks in and dysthy- SmithKline, Grunenthal GmbH, Janssen Pharma- mia. Psychopharmacol Bull 1997; 33: 101–103. ceutica, Jazz Pharmaceuticals Inc., Johnson & Johnson 8 Fava M, Rosenbaum JF, Pava JA, McCarthy MK, Steingard RJ, Pharmaceuticals, Knoll Pharmacuetical Company, Bouffides E. Anger attacks in unipolar depression, Part 1: clinical Lundbeck, MedAvante Inc., Merck, Neuronetics, correlates and response to fluoxetine treatment. Am J Psychiatry 1993; 150: 1158–1163. Novartis, Nutrition 21, Organon Inc., PamLab LLC, 9 Kessler RC, Merikangas KR. The National Comorbidity Survey Pfizer Inc., PharmaStar, Pharmavite, Roche, Sanofi- Replication (NCS-R): background and aims. Int J Methods Synthelabo, Sepracor, Solvay Pharmaceuticals Inc., Psychiatr Res 2004; 13: 60–68. Somaxon, Somerset Pharmaceuticals and Wyeth- 10 Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Ayerst Laboratories. He has been on speaker bureaus Eshleman S et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the for AstraZeneca, Boehringer Ingelheim, Bristol-Myers National Comorbidity Survey. Arch Gen Psychiatry 1994; 51: 8–19. Squibb Company, Cephalon, Eli Lilly and Company, 11 First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Forest Pharmaceuticals Inc., GlaxoSmithKline, Novar- Interview for DSM-IV-TR Axis I Disorders, Research Version, Non- tis, Organon Inc., Pfizer Inc., PharmaStar and Patient Edition (SCID-I/NP). Biometrics Research, New York State Psychiatric Institute: New York, NY, 2002. Wyeth-Ayerst Laboratories. He holds equity in Com- 12 Kessler RC, Berglund P, Chiu WT, Demler O, Heeringa S, Hiripi E pellis and MedAvante. Dr Kessler has been a et al. The US National Comorbidity Survey Replication (NCS-R): consultant for GlaxoSmithKline Inc., Pfizer Inc., design and field procedures. Int J Methods Psychiatr Res 2004; 13: Wyeth-Ayerst, Sanofi-Aventis, Kaiser Permanente 69–92. and Shire Pharmaceuticals; has served on advisory 13 Kessler RC, Ustun TB. The World Mental Health (WMH) Survey Initiative Version of the World Health Organization (WHO) boards for Eli Lilly and Company and Wyeth-Ayerst; Composite International Diagnostic Interview (CIDI). Int J Methods and has had research support for his epidemiological Psychiatr Res 2004; 13: 93–121. studies from Bristol-Myers Squibb, Eli Lilly and 14 American Psychiatric Association. Diagnostic and Statistical Company, GlaxoSmithKline, Johnson & Johnson Manual of Mental Disorders (DSM-IV), 4th edn. American Psychiatric Association: Washington, DC, 1994. Pharmaceuticals, Ortho-McNeil Pharmaceuticals 15 AJ, Carmody T, Reimitz PE. The inventory of depressive Inc., Pfizer Inc. and Sanofi-Aventis. Dr Rush has symptomatology (IDS): Clinician (IDS-C) and self-report (IDS-SR) provided scientific consultation to or served on ratings of depressive symptoms. Int J Methods Psychiatr Res 2000; Advisory Boards for Advanced Neuromodulation 9: 45–59. Systems Inc., AstraZeneca, Best Practice Project 16 Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of Depressive Symptomatology (IDS): psychometric Management Inc., Bristol-Myers Squibb Company/ properties. Psychol Med 1996; 26: 477–486. Otsuka, Cyberonics Inc., Forest Pharmaceuticals, 17 Hamilton M. A rating scale for depression. J Neurol Neurosurg GlaxoSmithKline, Jazz Pharmaceuticals, Eli Lilly Psychiatry 1960; 23: 56–62. and Company, Magellan Health Services, Merck & 18 Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein Co. Inc., Neuronetics, Novartis, Ono Pharmaceutical, DN et al. The 16-Item Quick Inventory of Depressive Symptoma- tology (QIDS), clinician rating (QIDS-C), and self-report (QIDS- Organon Inc., Pamlab Pfizer Inc. and Transcept SR): a psychometric evaluation in patients with chronic major Pharmaceuticals. He has received royalties from depression. Biol Psychiatry 2003; 54: 573–583. Guilford Press and Healthcare Technology Systems 19 Leon AC, Olfson M, Portera L, Farber L, Sheehan DV. Assessing and research/grant support from the National Insti- psychiatric impairment in primary care with the Sheehan Disability Scale. Int J Psychiatry Med 1997; 27: 93–105. tute of Mental Health and the Stanley Medical 20 SAS Institute. SAS 9.1.3 Help and Documentation. SAS Publish- Research Institute; has been on speaker bureaus for ing: Cary, NC, 2002. Cyberonics Inc., Forest Pharmaceuticals Inc. and 21 Wolter KM. Introduction to Variance Estimation. Springer-Verlag: GlaxoSmithKline; and owned stock in Pfizer Inc. New York, 1985. 22 Research Triangle Institute. SUDAAN: Professional Software for Survey Data Analysis. 8.01 edn. Research Triangle Institute: Research Triangle Park, NC, 2002. References 23 Fava GA, Pilowsky I, Pierfederici A, Bernardi M, Pathak D. Depressive symptoms and abnormal illness behavior in general 1 Crowe M, Ward N, Dunnachie B, Roberts M. Characteristics of hospital patients. Gen Hosp Psychiatry 1982; 4: 171–178. adolescent depression. Int J Ment Health Nurs 2006; 15: 10–18. 24 Merikangas KR, Akiskal HS, J, Greenberg PE, Hirschfeld 2 Ryan ND, Puig-Antich J, Ambrosini P, Rabinovich H, Robinson D, RM, Petukhova M et al. Lifetime and 12-month prevalence of Nelson B et al. The clinical picture of major depression in children bipolar spectrum disorder in the National Comorbidity Survey and adolescents. Arch Gen Psychiatry 1987; 44: 854–861. replication. Arch Gen Psychiatry 2007; 64: 543–552.

Molecular Psychiatry Irritability as a symptom of MDD M Fava et al 867 25 Benazzi F. Major depressive disorder with anger: a bipolar comorbid disorders characterize outpatients with anxious major spectrum disorder? Psychother Psychosom 2003; 72: 300–306. depressive disorder: a replication and extension. Can J Psychiatry 26 Akiskal HS. The emergence of the bipolar spectrum:validation 2006; 51: 823–835. along clinical-epidemiologic and familial-genetic lines. 34 Swanson JW, Holzer III CE, Ganju VK, Jono RT. Violence and Psychopharmacol Bull 2007; 40: 99–115. psychiatric disorder in the community: evidence from the 27 Bauer M, Pfennig A. Epidemiology of bipolar disorders. Epilepsia Epidemiologic Catchment Area surveys. Hosp Community Psy- 2005; 46(Suppl 4): 8–13. chiatry 1990; 41: 761–770. 28 Benazzi F. Possible bipolar nature of irritability in major depressive 35 Sher L, Oquendo MA, Li S, Ellis S, Brodsky BS, Malone KM et al. disorder. JClinPsychiatry2005; 66: 1072; author reply 1073. Prolactin response to fenfluramine administration in patients with 29 Benazzi F, Akiskal H. Irritable-hostile depression: further valida- unipolar and bipolar depression and healthy controls. Psycho- tion as a bipolar depressive mixed state. J Affect Disord 2005; 84: neuroendocrinology 2003; 28: 559–573. 197–207. 36 Davis LL, Clark DM, Kramer GL, Moeller FG, Petty F. D- 30 Alpert JE, Petersen T, Roffi PA, Papakostas GI, Freed R, Smith MM fenfluramine challenge in posttraumatic stress disorder. Biol et al. Behavioral and emotional disturbances in the offspring of Psychiatry 1999; 45: 928–930. depressed parents with anger attacks. Psychother Psychosom 37 New AS, Trestman RF, Mitropoulou V, Goodman M, Koenigsberg 2003; 72: 102–106. HH, Silverman J et al. Low prolactin response to fenfluramine in 31 Perris C. The distinction between unipolar and bipolar mood impulsive . J Psychiatr Res 2004; 38: 223–230. disorders. A 25-years perspective. Encephale 1992; 18(Spec no. 1): 38 Fava M, Vuolo RD, Wright EC, Nierenberg AA, Alpert JE, 9–13. Rosenbaum JF. Fenfluramine challenge in unipolar depression 32 Yoshimasu K, Sugahara H, Tokunaga S, Akamine M, Kondo T, with and without anger attacks. Psychiatry Res 2000; 94: 9–18. Fujisawa K et al. Gender differences in psychiatric symptoms 39 Dougherty DD, Rauch SL, Deckersbach T, Marci C, Loh R, Shin LM related to suicidal ideation in Japanese patients with depression. et al. Ventromedial prefrontal cortex and amygdala dysfunction Psychiatry Clin Neurosci 2006; 60: 563–569. during an anger induction positron emission tomography study in 33 Fava M, Rush AJ, Alpert JE, Carmin CN, Balasubramani GK, patients with major depressive disorder with anger attacks. Arch Wisniewski SR et al. What clinical and symptom features and Gen Psychiatry 2004; 61: 795–804.

Molecular Psychiatry